High-water marks · a two-wall model

A higher line on a wall does not have to mean higher water

We read a flood mark from today's pavement, but a pavement is not a contour line, and the street may have risen since the flood. Change what you measure the two lines against.

Without JavaScript you see example 1 in the “From today's pavement” view. All three readings of both examples are in the tables below (explanation).

Result of comparing the two lines

Difference between the lines · above today's pavement

0.90m

B is higher

Apparent paradox: from today's pavement B is higher (by 0.90 m), but in the common Z system A is higher (by 1.50 m).

An equal or different elevation is not yet proof of a single water surface: a river has a gradient, a mark may have been moved, and the data in the model are samples.

The same two lines, three readings at once

Model assumptions (an example, not a measurement): wall A — mark 1.80 m above today's pavement, ground Z = 12.40 m, street raised by 0 cm. Wall B — mark 2.70 m, ground Z = 10.00 m, street raised by 0 cm.

Height of marks A and B in three readings, and which is higher
Reading	Wall A	Wall B	Higher
Above today's pavementhin view	1.80 m	2.70 m	B +0.90 m
Above the cobbles of the day of the floodh + p, where p is the rise of the street	1.80 m	2.70 m	B +0.90 m
Elevation Z of the mark in a conventional systemZ = Z of the ground + h	14.20 m	12.70 m	A +1.50 m

Measure the lines against

Vertical axis: metres above today's pavement (zero = the pavement beside each wall). A schematic of the model, not a map or a measurement.

water mark (blue line and its level)
today's pavement
cobbles of the flood day and accumulated layers
ground under the wall

What is equal here: both of today's pavements lie on one zero line. What has been hidden: the ground under A lies 2.40 m higher than under B. The door (2.10 m) and the person (1.75 m) are a conventional scale.

Change wall B

Ground under wall B (elevation Z) 10.0 m

The street at B has risen since the flood by 0 cm

I do not know the ground elevation under B

The sliders also work with the keyboard: arrow keys, Home, End. The rise of the street is a hypothesis to try out, not a correction based on data.

One line, three readings

Above today's pavement (h): This is what a passer-by sees. It depends on how high the pavement lies today, so two walls on different ground do not give comparable numbers.
Above the cobbles of the day of the flood (h + p): If the street has risen by p since the flood, the cobbles of the time lay lower and the water reached higher above them. The value of p is a hypothesis in the model: to know it, one would have to study the accumulated layers or documents.
Elevation of the mark (Z = Z of the ground + h): Adding makes sense only if both ground elevations are in the same system and refer to the right moment. The system in the model is conventional; it is not a conversion to m a.s.l. A change of p alone does not move the mark, so it does not change Z.

Consistency of the numbers: Z − (Z of the ground − p) = h + p. That is why the street slider changes only the second row of the table (h + p). The first row (h) and the third (Z) stay the same, because the mark has not moved from the wall.

Two examples in numbers model

Values of both examples in three readings
Example	Reading	A	B	Higher
1 · A: h 1.80, ground 12.40 · B: h 2.70, ground 10.00, p 0	above the pavement	1.80 m	2.70 m	B +0.90 m
1 · as above	above the cobbles of the flood day	1.80 m	2.70 m	B +0.90 m
1 · as above	elevation Z	14.20 m	12.70 m	A +1.50 m
2 · A: h 1.90, ground 11.50 · B: h 1.50, ground 11.90, p 40 cm	above the pavement	1.90 m	1.50 m	A +0.40 m
2 · as above	above the cobbles of the flood day	1.90 m	1.90 m	equal
2 · as above	elevation Z	13.40 m	13.40 m	equal

A benchmark is a different kind of mark. A metal geodetic benchmark on a wall serves the national height control network and does not document floods. The inscription, the date of the event and the time the plaque was made are three separate pieces of information. A sample photograph of a benchmark from Toruń is only linked in the sources, not embedded.

What this model does not say

Walls A and B are invented. The heights, ground elevations and street rises are sample values, not measurements of any building.
An equal elevation of two marks proves neither a single flood nor a level water surface. A river has a gradient, a mark may have been moved or recreated, and the ground elevation may have changed.
An unknown elevation is not zero. When you tick “I do not know the ground elevation under B”, wall B gets no place on the Z axis and no ranking arises in the common system.
The 2.10 m door and the 1.75 m person are a conventional yardstick. In the “from the cobbles of the flood day” view the door is left out, because we do not know where its threshold lay.
This page does not assess flood risk or the safety of any building, and it is not a land survey document.

To check before using real marks: whether the mark refers to the same event, whether it was moved, how the street level changed, and whether the elevations are given in the same system and from the same date.

What changed compared with version 2 V3

The fields for entering your own heights, names and units were removed. Without elevations in a single system, the numbers entered would give a false sense of precision. They were replaced by two examples and two sliders for wall B.
Saving to browser storage was removed. There is nothing in it to keep, and the page neither reads nor deletes the old key atlas-v2-131.
The apparent-paradox detector was kept, but it works on a single documented set of model data, visible at once in all three readings.
The formula Z = Z of the ground + h, the units in metres and the block on aligning when an elevation is missing were kept.
The section “11 April 1829. One event, different walls” was removed from the page content. The date and locations (Baszta Łabędź, Brama Nizinna) remain only in the description of the UMK source, with the status “to be checked”, because the two-wall model is a sample and does not concern specific marks.
The three pitfall cards from version 2 were shortened to paragraphs in the explanation and in the list of the model's limits; there are no separate cards any more.

Sources and their status

The links come from version 2. They were not checked again in this version, so they are marked to be checked. The model's numbers do not come from them.

Wody Polskie, opis bazy WORP [Polish Waters, description of the WORP database], 4 December 2018, p. 23: definition of the fields of the ZNAKI_WW layer (date, elevation, height above ground). to be checked
UMK, Znaki wielkiej wody [High-water marks] (presentation): according to version 2 it gives the date 11 April 1829 and locations on the Baszta Łabędź and the Brama Nizinna in Gdańsk. The current state of the marks needs checking. to be checked
Wikimedia Commons: Torun_Ratusz_reper.jpg, photo by Pko, 8 March 2008, CC BY-SA 4.0. A geodetic benchmark, not a flood mark. The photograph is not embedded in this page.

The two-wall schematic is the author's own drawing of the model (HTML, CSS and a simple SVG silhouette), with no photographs and no external resources. The page works offline.